Progress 04/25/16 to 04/18/21
Outputs
PROGRESS REPORT Objectives (from AD-416): 1. Determining the safety of low and alternative salt fermentations, produced nationally and internationally. 2. Develop predictive models for 5-log reduction times for pathogenic Escherichia coli in fermented and acidified vegetable products. 3. Enhance buffer capacity models for predicting pH changes in acidified foods with low acid ingredients. Approach (from AD-416): The experimental approaches that will be use to achieve the objectives will include mathematical modeling, molecular ecology studies, and biochemical analysis of fermentation brines. Specifically, for Objective 1, to determine the effects of salts on pathogen reduction in fermentations, growth and death of bacterial pathogen cocktails (strain mixtures) will be measured in fermentations by conventional bacterial plating methods using automated plating equipment. Log reduction times for pathogens will be calculated using linear or nonlinear (Weibull) models. Biochemical analysis for salts, organic acids and sugars, will be done by titration (for salts), and high performance liquid chromatography (for acids and sugars). A matrix of salt types and concentrations will be tested to determine how salt effects pathogen die-off. For Objective 2, mathematical modeling approaches to determine the reduction in pathogen populations during fermentation will utilize non-linear systems of ordinary differential equations (rate equations) using Matlab computer software. In addition computer simulation models will be developed using the C++ programming language. Data for these models will be obtained from the experiments in Objective 1. Model results will be compared to data generated under a variety of conditions to determine if the models accurately describe the data. To accomplish Objective 3, predicting pH of buffered acidified foods with low acid additives, mathematical models will be based on published ionic equilibria equations for buffered acid and base solutions. Novel methods for numerical solutions to these equations will be implemented with Matlab software. An automated titrator will be used to confirm predicted buffer capacity curve data. To fit data to the models, several optimization algorithms will be used from the Matlab Optimization Toolkit, or independently programmed in Matlab or C++. The knowledge gained will be used to help processors and regulatory agencies assess and assure the safety of acidified and fermented food products. This is the final report for the Project 6070-41420-008-00D terminated in April 2021,which has been replaced by new Project 6070-41420-009-00D. Buffer capacity modeling. ARS scientists at Raleigh, North Carolina suggests buffer capacity is a critical factor for determining how pH may change with the addition of ingredients to acid or acidified foods, which are made safe by low pH. While models have been published in the scientific literature for predicting pH in water-based solutions, ARS scientists at Raleigh, North Carolina have developed a method for identifying buffers in undefined food ingredient solutions and using the data to predict how pH may change with added food ingredients for acidic foods. The models were used by ARS scientists at Raleigh, North Carolina for determining the buffer capacity of typical ingredients in these food products, and calculating how these ingredients may change pH, a vital safety factor for foods primarily protected from disease causing bacteria by pH. ARS scientists at Raleigh, North Carolina have developed mathematical models and methods for the analysis of food ingredients to allow pH predictions in a final product with mixed ingredients and acids typical of salad dressing products. The models have been shown to accurately predict pH acid-base solutions and can predict relative amount of buffering present in different food ingredients. These models have broad applicability with many acid or acidified foods, and aid producers and regulatory agencies in determining the safety of acidified foods products. Bacterial competition in vegetable fermentations. The acid resistance of pathogenic Escherichia coli strains likely explains the pathogen⿿s ability to transit the human stomach, which makes these organisms a concern in acid or acidified foods. ARS scientists at Raleigh, North Carolina screened multiple E. coli strains to assess their acid resistance in simulated stomach acid and vegetable fermentation acids. These strains were separated into two groups that differed in acid resistance, measured their growth rates, and ability to survive in competition with lactic acid bacteria in model vegetable fermentations. Genome sequence data showed that one of the acid sensitive strains was missing two acid resistance genes. Unexpectedly, ARS scientists at Raleigh, North Carolina found that the acid sensitive strain was as competitive or more competitive than the acid resistant strains in model vegetable fermentations. These results indicate that the resistance to acids in laboratory tests may not be an accurate predictor of E. coli survival in vegetable fermentations and have triggered new research on efforts to understand how bacterial competition works in fermented foods. Viruses active against enteric bacteria in vegetable fermentations. Bacteriophage are viruses that specifically attack bacteria and are widely present in the natural environment and in food fermentations. The diversity and functional role that these phages play in infecting Enterobacteriaceae, which are spoilage and potentially pathogenic bacteria that can grow in the early stages of vegetable fermentations, was investigated. In samples from the first few days of a commercial vegetable fermentation 26 independent bacteriophage were isolated, along with 39 Enterobacteriaceae strains that were tested to see if the phage would infect them. Two-thirds of the phage were active against these cultures. This study by ARS scientists at Raleigh, North Carolina, combined with previous similar studies of sauerkraut fermentations, revealed the abundance and variety of phages infecting Enterobacteriaceae bacteria in the early stages of vegetable fermentations. The data show that both bacteriophage infection and inhibition by lactic acid both play a role in maintaining the safety of vegetable fermentations by reducing Enterobacteriaceae populations. Critical controls for fermented foods. Under the Food Safety Modernization Act, a business that manufactures fermented foods may be required to conduct a risk analysis and establish pertinent preventive controls. Retail food establishments operating under the Food and Drug Administration (FDA) Food Code must often seek a variance for manufacture of fermented foods and beverages. Developing food safety programs can be a challenge for small-scale producers with little access to training and resources, especially as manufacture of fermented products involves microbiologically complex systems that may not be effectively or appropriately managed by standard time-temperature controls. Working with university extension faculty, the science behind traditional vegetable fermentation processes, e.g., cabbage, cucumbers and peppers, was reviewed by ARS scientists at Raleigh, North Carolina to identify relevant hazards based on intrinsic and extrinsic factors inherent in the fermentation systems that influence microbial survival. It was determined by ARS scientists at Raleigh, North Carolina that one Critical Control Point in the manufacture of traditionally fermented vegetable products, namely a steady and sustained pH decline to < 4.6, was important for food safety. Additional Control Points at key steps, i.e., vegetable preparation and salt addition; fermentation time and temperature; refrigerated storage; and/or packaging for shelf stability were also identified by ARS scientists at Raleigh, North Carolina, and an example for safe manufacture of fermented vegetables was developed, based on the fermentation of kimchi. Microbiota in vegetable fermentations with varying salt levels. The influence of salt type used in cover brines on the microbiota of laboratory and commercial scale cucumber fermentations was investigated by ARS scientists at Raleigh, North Carolina. Laboratory fermentation cover brines with calcium chloride (low salt brining technology) induced faster microbial growth as compared to cover brines with no salt or 6% sodium chloride typical of traditional commercial fermentations. In the initial days of fermentation Enterobacteriaceae such as Citrobacter and Enterobacter was favored in fermentations brined with sodium chloride, in which it took longer for the lactic acid bacteria to grow compared to other salt conditions. Lactobacilli dominated all fermentation brines by the third day, regardless of salt type or content, and 80 to 88% of the population was composed of lactobacilli by the seventh day of fermentation, except in fermentations without salt, in which a mixed population of lactic acid bacteria were still prevalent. In general, the population of lactic acid bacteria found in commercial cucumber fermentations brined with 1.1% calcium chloride were similar to the bacteria found in laboratory fermentations. Understanding how salt affects the microbiota of commercial fermentations will facilitate optimization of low salt fermentation technology. Summary report for 2015-2021. Significant results were achieved by ARS scientists at Raleigh, North Carolina over the life of the project, which was completed in 2021. Results from research on Objective 1 have shown that fermentation salts (sodium and calcium chlorides) primarily enhance the growth of lactic acid bacteria and therefore are only indirectly inhibitory to vegetative bacterial pathogens such as Escherichia coli, Salmonella enterica and Listeria monocytogenes. The most acid resistant pathogen of concern (E. coli) was found to grow in the early stages of vegetable fermentations using a model system with no plant inhibitors of bacterial growth, even if the salt concentration was high (6%). However, these organisms rapidly died off as organic acid increased, and pH dropped. Surprisingly, data showed that one of the most acid resistant E. coli strains may not survive as long in fermentations as a more acid sensitive resistant strain. This was likely due to the affect the acid resistant E. coli strain had on buffering the pH of the medium, aiding growth of competing lactic acid bacteria. The published data shed new light on the complex interactions that occur during bacterial competition in fermented foods. Results from competitive growth modeling efforts for Objective 2 showed that a key factor, predicting pH in the fermentation medium, was needed for these models to be widely applicable to fermented vegetables. To address this research need, buffer models developed by ARS scientists at Raleigh, North Carolina for Objective 3 (see below) were adapted to a model vegetable fermentation system. Using in silico models, it was found that pH predictions based on the initial buffering of the medium were useful for predicting the extent of fermentation pH changes. Further development of this technology is an important objective in our subsequent food safety research plans. Buffer capacity (BC) models to predict the changes in pH due to low acid ingredients in acid and acidified foods were successfully developed for Objective 3. The modeling effort included the development of a series of algorithms that can be used to: 1) generate BC models from titration data; 2) model BC curves to generate data on buffers in a food ingredient that control pH; 3) Allow comparison of relative buffering of different food ingredients; 4) predict pH for a given set of ingredients. These algorithms were encoded using Matlab modeling software and made available for public download on the ARS software website. A stand-alone Windows 10 program that may be used for comparison of ingredient buffering was also developed and similarly published. Model results have been used by industry trade associations and made available to FDA to aid in the regulatory process of differentiating different types of acidic food products. Record of Any Impact of Maximized Teleworking Requirement: Data needed for mathematical models of bacterial competition for assessing the impact of mixed organic acids on the growth and survival of bacterial pathogens was not obtained by ARS scientists at Raleigh, North Carolina. A statistical design for these experiments was developed by ARS scientists at Raleigh, North Carolina, however, in collaboration with university statisticians. This research is planned for 2021 and has been incorporated into our subsequent NP108 project plan. Research on the use of hops as a natural antimicrobial under a CRADA agreement was suspended during the maximized telework period. Preliminary data had shown that selected hop compounds have promise for use as a natural alternative to sodium benzoate in acid and acidified foods. Efforts to renew the research project with the CRADA partner are under way. Initiation of research for a new project to investigate the effect of starch compounds on the survival of Listeria in selected salad dressing products did not occur in 2020. Preliminary data has shown that some starch compounds may prolong Listeria survival in these foods but may not prolong survival of other pathogens such as Escherichia coli. With changeover in personnel, this project is currently on hold, but may be resumed as funding and personnel become available. ACCOMPLISHMENTS 01 Determination of the buffer capacity of ingredients in acidic foods. ARS researchers in Raleigh, North Carolina used recently developed buffer capacity models to analyze food ingredients commonly used in salad dressings and related pickled vegetable products. This was done by combining classical physical chemistry concepts of pH modeling with modern numerical computing methods. For 24 food ingredients commonly used in acidic food products like salad dressings, including acids as well as flavoring ingredients and spices that have a pH close to neutral, a model of buffering was developed by ARS scientists at Raleigh, North Carolina. These models were then used by ARS scientists at Raleigh, North Carolina to estimate how each ingredient, if added to an acidic food product, would (or would not) change pH of the product. Compared to the acids typically added in these acidic foods (usually acetic acid) most other salad dressing ingredients tested had very little influence on pH change at concentrations typically used for commercial products. This was found to be due to very weak buffering activity. For acidic food products where pH is a critical control factor for food safety, these data can be used to quantify safe concentration limits. The results may be useful to both manufacturers and regulatory agencies for defining safe production practices and to assure product stability. 02 Determining how bacterial competition influences the rate of reduction of pathogenic bacteria in vegetable fermentations with different salt concentrations. ARS scientists at Raleigh, North Carolina believe the effects of microbial competition on the survival of pathogenic bacteria in vegetable fermentations has previously been demonstrated, but quantitative data on pathogen die-off is lacking. To investigate the effects of fermentation conditions on the survival of acid resistant Escherichia coli strains, ARS scientists at Raleigh, North Carolina measured the die-off of different pathogenic E. coli strains in laboratory vegetable fermentations. Unexpectedly one of the most acid resistant strains of E. coli did not survive as long as a less acid resistant strain in competition with lactic acid bacteria (LAB). The data indicated that competition with LAB is dependent on complex factors that include buffering of the fermentation medium. This buffering can alter expected pH changes and influence all bacteria in the fermentation (including E. coli). Methods to measure buffering have been developed by ARS scientists at Raleigh, North Carolina for use in future studies. These data aid both manufacturers and regulatory agencies concerned with assuring safety of fermented vegetable foods. 03 Enhanced safety of refrigerated cucumber pickles. A brief blanching procedure was developed by ARS scientists at Raleigh, North Carolina to improve the safety of refrigerated cucumber pickles. The traditional manufacturing process for these products has limited efficacy against pathogenic bacteria which may be on cucumbers from environmental sources. The brief blanching procedure for raw cucumbers was investigated by ARS scientists at Raleigh, North Carolina to determine the reduction in bacterial cell counts as well as the effect of blanching on finished product quality. The data were used by ARS scientists at Raleigh, North Carolina to model pathogen kill at varying depths within the cucumber for an optimized procedure of 90 seconds of blanching in 80°C (176°F) water. The modeling was done in collaboration with North Carolina State University research engineers. The resulting model can show the predicted death of acid resistant strains of Escherichia coli, which is the most acid resistant pathogen of concern for these products. Unexpectedly, the data also showed that the blanching procedure was effective in improving some quality aspects of refrigerated pickle products made with cucumbers subjected to the blanch treatment. These data have been presented to industry stakeholders and may be of general use to the pickled vegetable industry for improving refrigerated product safety.
Impacts
(N/A)
Publications
|
Progress 10/01/19 to 09/30/20
Outputs
Progress Report Objectives (from AD-416): 1. Determining the safety of low and alternative salt fermentations, produced nationally and internationally. 2. Develop predictive models for 5-log reduction times for pathogenic Escherichia coli in fermented and acidified vegetable products. 3. Enhance buffer capacity models for predicting pH changes in acidified foods with low acid ingredients. Approach (from AD-416): The experimental approaches that will be use to achieve the objectives will include mathematical modeling, molecular ecology studies, and biochemical analysis of fermentation brines. Specifically, for Objective 1, to determine the effects of salts on pathogen reduction in fermentations, growth and death of bacterial pathogen cocktails (strain mixtures) will be measured in fermentations by conventional bacterial plating methods using automated plating equipment. Log reduction times for pathogens will be calculated using linear or nonlinear (Weibull) models. Biochemical analysis for salts, organic acids and sugars, will be done by titration (for salts), and high performance liquid chromatography (for acids and sugars). A matrix of salt types and concentrations will be tested to determine how salt effects pathogen die-off. For Objective 2, mathematical modeling approaches to determine the reduction in pathogen populations during fermentation will utilize non-linear systems of ordinary differential equations (rate equations) using Matlab computer software. In addition computer simulation models will be developed using the C++ programming language. Data for these models will be obtained from the experiments in Objective 1. Model results will be compared to data generated under a variety of conditions to determine if the models accurately describe the data. To accomplish Objective 3, predicting pH of buffered acidified foods with low acid additives, mathematical models will be based on published ionic equilibria equations for buffered acid and base solutions. Novel methods for numerical solutions to these equations will be implemented with Matlab software. An automated titrator will be used to confirm predicted buffer capacity curve data. To fit data to the models, several optimization algorithms will be used from the Matlab Optimization Toolkit, or independently programmed in Matlab or C++. The knowledge gained will be used to help processors and regulatory agencies assess and assure the safety of acidified and fermented food products. Fermented vegetable foods have become increasingly popular, but there has been little information available to aid manufacturers of fermented vegetable products in developing food safety programs that can meet Food Safety Modernization Act (FSMA) and Food and Drug Administration (FDA) Food Code requirements. Under FSMA regulations, a business that manufactures fermented foods may be required to conduct a risk analysis and establish pertinent preventive controls. To fill this knowledge gap, ARS researchers at Raleigh, North Carolina, identified pH as the sole critical control for the manufacture of traditionally fermented vegetable products. This includes a steady and sustained pH decline to < 4.6 for prevention of spore outgrowth and botulism. ARS researchers at Raleigh, North Carolina, linked 5-log reduction times to pH for the principal microbial hazards including Escherichia coli, Salmonella enterica, and Listeria monocytogenes. Additional control points were identified for key processing steps, and included vegetable preparation (cleaning and washing) and salt addition (= 2% is recommended, although lower salt concentrations will work); fermentation time and temperature (e.g. typically 65-72°F for three or more weeks for cabbage fermentations); and refrigerated storage and/or hermetically sealed packaging for shelf- stability. While the parameters for the above control points may differ depending on fermentation type, the controls identified will be beneficial to aid producers in meeting FDA regulatory requirements and establishing production controls for fermented foods, including cucumber pickles, sauerkraut and others. While it is well known that competition between lactic acid bacteria and bacterial pathogens in vegetable fermentations results in safe fermented products, a mechanistic understanding of bacterial competition is lacking. To help fill this void, ARS researchers at Raleigh, North Carolina, investigated how acid resistance characteristics of disease-causing Escherichia coli (E. coli) strains influenced survival of this pathogen in vegetable fermentations. Surprisingly, ARS researchers at Raleigh, North Carolina, found that the most acid resistant E. coli strains did not survive as well as more acid sensitive strains under some laboratory fermentation conditions. These results are important because they indicate that there are chemical and environmental factors driving changes in microbial ecology during fermentations that are poorly understood and could directly affect fermentation safety. Specifically, the data revealed that metabolic reactions of E. coli that result in pH buffering may be important for controlling E. coli survival in competition with lactic acid bacteria. These results will help guide future research to further define the conditions leading to the safe manufacture of a variety of fermented vegetable products. A cell-based model for the competitive growth of lactic acid bacteria and bacterial pathogens in vegetable fermentations has been developed in Matlab. This model has been developed by ARS researchers at Raleigh, North Carolina, as a computer simulation based on mechanistic principles for cell growth and division, rather than traditional growth modeling based on cell population growth and death rates. A novel feature of the model is the prediction of pH based on acid production by the cells undergoing fermentation, taking into account the buffering of the fermentation medium itself. An advantage of the cell-based mechanistic modeling is that bacterial competition and pH changes in fermentation systems are not dependent on the specific conditions of the experiments or defined by fitted curves. Using curve fitting methods for these variables limits the utility of the models to only the system studied. Further development of these models may lead to practical use of pH as a measure of fermentation safety because these models link pH directly to acid concentrations, which has been difficult for mixed acid fermentations. The pH of most acid food products depends on undefined and complex buffering of all ingredients but is critically important for regulatory purposes and food safety. ARS researchers at Raleigh, North Carolina, have developed a new method for modeling the pH of complex food ingredients. ARS researchers at Raleigh, North Carolina, used the method for defining the buffering and pH of individual ingredients and ingredient mixtures in salad dressing products. Ingredients of salad dressings were titrated individually and in combination at concentrations typical of dressing products. Buffer data were then used to predict pH. The research showed that most ingredients in salad dressings had little buffering compared to the vinegar typically used in dressing formulations, meaning that very little pH change would occur due to adding the ingredient to the dressing. ARS researchers at Raleigh, North Carolina, found that sugars showed significant buffering at high pH values due to very weakly acidic hydroxyl groups on sugar molecules. These chemical groups would not affect product pH under conditions typical of foods. Application of buffer models and data for dressing ingredients may help manufacturers with new product formulation and determining pH stability and safety of dressing products. This work will also aid regulatory agencies in assessing the impact of ingredients in salad dressings on final product pH for regulatory purposes. Accomplishments 01 Development of buffer capacity models for acid and acidified foods. Acid foods are primarily composed of acidic food ingredients but often have small amounts of spices and other high pH ingredients to improve flavor, texture or other properties of the final products. Acidified foods, on the other hand are primarily composed mostly of high pH ingredients (like cucumbers) that are made acidic by adding vinegar. Defining these two product types, which are regulated differently for safety considerations by FDA, has been difficult. ARS researchers at Raleigh, North Carolina, have developed buffer models and modeling methods that can be used to define the pH impact of all the ingredients in different types of acid or acidified foods. The method works even if the chemical compositions of the food ingredients are not known or not chemically defined. The buffer capacity models have been adapted by the salad dressing industry and by FDA to help differentiate between acid and acidified foods based on pH changes with ingredient addition. These models represent a novel scientific advance with broad application to the safety of acid and acidified foods, including: 1) predicting pH of products based on ingredient mixtures; 2) estimating pH stability; and 3) predicting pH changes in vegetable and other food fermentations as acids accumulate.
Impacts
(N/A)
Publications
- Longtin, M., Price, R.E., Mishra, R., Breidt, F. 2020. Modeling the buffer capacity of ingredients in salad dressing products. Journal of Food Science. 85(4):910-917.
- Price, R.E., Longtin, M., Conley Payton, S., Osborne, J.A., Johanningsmeier, S.D., Bitzer, D., Breidt, F. 2020. Modeling buffer capacity and pH in acid and acidified foods. Journal of Food Science. 85(4) :918-925.
- Jones, C.M., Price, R.E., Breidt, F. 2020. Escherichia coli O157:H7 stationary phase acid resistance and assessment of survival in a model vegetable fermentation system. Journal of Food Protection. 83(5):745-753.
- Snyder, A., Breidt, F., Andress, E.L., Ingham, B.H. 2020. Manufacture of traditionally fermented vegetable products: Best practice for small businesses and retail food establishments. Food Protection Trends. 40(4) :251-263.
- Lu, Z., Perez Diaz, I.M., Hayes, J., Breidt, F. 2020. Bacteriophages infecting gram-negative bacteria in a commercial cucumber fermentation. Frontiers in Microbiology. 11:1306.
|
Progress 10/01/18 to 09/30/19
Outputs
Progress Report Objectives (from AD-416): 1. Determining the safety of low and alternative salt fermentations, produced nationally and internationally. 2. Develop predictive models for 5-log reduction times for pathogenic Escherichia coli in fermented and acidified vegetable products. 3. Enhance buffer capacity models for predicting pH changes in acidified foods with low acid ingredients. Approach (from AD-416): The experimental approaches that will be use to achieve the objectives will include mathematical modeling, molecular ecology studies, and biochemical analysis of fermentation brines. Specifically, for Objective 1, to determine the effects of salts on pathogen reduction in fermentations, growth and death of bacterial pathogen cocktails (strain mixtures) will be measured in fermentations by conventional bacterial plating methods using automated plating equipment. Log reduction times for pathogens will be calculated using linear or nonlinear (Weibull) models. Biochemical analysis for salts, organic acids and sugars, will be done by titration (for salts), and high performance liquid chromatography (for acids and sugars). A matrix of salt types and concentrations will be tested to determine how salt effects pathogen die-off. For Objective 2, mathematical modeling approaches to determine the reduction in pathogen populations during fermentation will utilize non-linear systems of ordinary differential equations (rate equations) using Matlab computer software. In addition computer simulation models will be developed using the C++ programming language. Data for these models will be obtained from the experiments in Objective 1. Model results will be compared to data generated under a variety of conditions to determine if the models accurately describe the data. To accomplish Objective 3, predicting pH of buffered acidified foods with low acid additives, mathematical models will be based on published ionic equilibria equations for buffered acid and base solutions. Novel methods for numerical solutions to these equations will be implemented with Matlab software. An automated titrator will be used to confirm predicted buffer capacity curve data. To fit data to the models, several optimization algorithms will be used from the Matlab Optimization Toolkit, or independently programmed in Matlab or C++. The knowledge gained will be used to help processors and regulatory agencies assess and assure the safety of acidified and fermented food products. ARS researchers have data for delta psi measurements of acid stressed pathogenic Escherichia coli (STEC), however, continued progress on the internal pH, delta psi and internal salts/metabolites of STEC strains has been suspended because of the loss of a scintillation counter at our location that is needed for quantifying radionuclide activity in samples. Related work on growth and survival of STEC in model vegetable fermentation brines was completed, showing the growth and death rates of STEC strains in both calcium chloride and traditional sodium chloride fermentation brines with a variety of salt levels selectively favored natural fermentative lactic acid bacteria as salt increased. Data from commercial fermentations showed that higher salt levels favored maximum growth levels of normal fermentative bacteria over potential pathogens, although pre-acidification of brines prior to fermentation with acetic acid could reduce survival of enteric (potentially pathogenic) bacteria. This is important because low salt calcium chloride fermentation technology is being developed to reduce waste salt in commercial vegetable fermentations. A computer simulation model to be used for the competitive growth of bacteria was developed using C++ programming language. This mechanistic model has parameters previous determined to control growth at a cell- based level, including per-cell rates of sugar utilization, and the amount of sugar needed for cell division. A multi-threaded program was implemented on a North Carolina State University High Performance Computing (HPC) System server, in collaboration with NC State HPC personnel. The model has been transferred to the ARS CERES computer system via Scinet for continued development by our research group. High throughput processing is required for predicting bacterial growth for up to 1 billion cells. Parameters for acid inhibition of cell death depend on the evolving pH in fermentation brines, and therefore required development of buffer capacity models (BC models, see below) for fermentation brines. Vegetable broths were fermented to various stages and titrated to determine BC by titration and then estimating pH by the BC model. Preliminary data has shown that systematic changes in BC of the fermentations may allow pH prediction over time, based on production of fermentation acids and cell growth. This has been a major stumbling block with previous models because only empirical methods were available for pH prediction. Buffer capacity (BC) models were developed to predict the stability of pH changes in acid and acidified food systems. The models were validated with food industry samples. These models consist of two components: first is the determination of a matrix of weak acids and bases present in foods with chemically undefined food ingredients, and second is the pH prediction from these matrices. The BC model process that has been developed requires both laboratory methods and computer models. The first step was titration of food ingredients between pH 2 and pH 12 using an automated titrator. While not designed for generating BC data, the dynamic dosing feature of the titrator was used to generate smooth curves for both acid and base titrations. Data from the titrator was imported into the Matlab computer program using custom (Python) computer code. Once in Matlab, a novel curve fitting method using trigonometric regression was used to analyze the data. This was followed by an optimization algorithm to determine a matrix of buffer values consisting of concentration and equilibrium constant values for weak acids and bases in solution. Finally, pH prediction was done based on the buffer values. These modeling steps were implemented using a series of custom programs that are controlled by a Matlab ⿿live-script file. The model has been successfully validated and used to examine salad dressing products, as well as fermentation brines from vegetable fermentations. Accomplishments 01 Determining the effects of Sodium Chloride or Calcium Chloride concentration on the growth and death of bacterial pathogens in vegetable fermentations. Salt concentration has long been considered an important factor for the quality of fermented vegetable products, but the role of salts in bacterial growth and death during vegetable fermentation remains unclear. ARS scientists in Raleigh, North Carolina, compared the effects of sodium chloride and calcium chloride used in commercial cucumber fermentations on the growth and death of the normal fermentation lactic acid bacteria (LAB) and pathogenic bacterial strains that could be in the brines. The data showed that low salt concentrations had a stimulatory effect on the growth rates of bacteria compared with a no-salt control, but higher salt concentrations decreased growth rates for pathogens; to a lesser extent, LAB growth rates were also reduced. However, no consistent pattern was observed when comparing pathogen death rates with salt type or concentration. For vegetable fermentation safety concerns, the results suggest that an important effect of salt addition is enhancement of the growth of LAB compared to pathogenic Escherichia coli strains. 02 Effects of brine acidification on cucumber fermentation bacteria in calcium or sodium chloride brines. Commercial fermentation for bulk preservation of cucumbers typically relies on natural microbiota and high salt sodium chloride (NaCl) brines. An alternative process utilizing low salt calcium chloride brines was previously developed to eliminate NaCl from fermentation brines for reduced environmental impact. ARS researchers in Raleigh, North Carolina, found that potentially pathogenic enteric bacteria survived the longer in the cucumbers in 6% NaCl brines compared to calcium brines with no initial brine acidification, likely due to slower fermentation under these conditions. However, the addition of acetic acid as a pre-treatment to fermentation brines significantly reduced the survival of these bacteria in both calcium and sodium salt treatments. These data show that pre-acidification of fermentation brines can improve fermentation safety. 03 Development of hot-fill pasteurization of cucumber pickle spears as an alternative to tunnel pasteurization. For commercial production of acidified vegetable products, a tunnel pasteurizer is typically used for thermal processes. To help reduce energy costs and use of water, ARS researchers in Raleigh, North Carolina, developed a hot-fill method for pasteurization of cucumber pickle spears in 24 oz pickle jars. The method required refilling jars multiple times with a hot brine (around 175oF). The data showed that for cucumber spears a hot fill method could achieve or exceed temperatures typically used for commercial pasteurization of pickle by most manufacturers. These conditions exceed published values needed for the required reduction of bacterial pathogens in acid and acidified vegetable products and were sufficient to meet typical industry processing conditions to assure good quality texture and sensory properties. Although further development of processing equipment may be needed for inverting and refilling jars, the in-jar pasteurization process has potential application for cucumber spears and related products and may be used to save on the water usage and costs of currently used tunnel pasteurizers.
Impacts
(N/A)
Publications
- Breidt, F., Andress, E., Ingham, B. 2018. Recommendations for designing and conducting cold-fill hold challenge studies for acidified food products. Food Protection Trends. 38(5):322-328.
- Dupree, D., Price, R.E., Burgess, B., Andress, E., Breidt, F. 2019. Effects of sodium chloride or calcium chloride concentration on the growth and survival Escherichia coli O157:H7 in model vegetable fermentations. Journal of Food Protection. 82(4):570-578.
- Yavuz, N., Foster, L., Sharma, T., Patel, K., Stoforos, G., Sandeep, K.P., Planitkar, P., Breidt, F. 2019. Hot-fill pasteurization of cucumber pickle spears: An alternative to tunnel pasteurization. Food Protection Trends. 38(4):258-265.
- McMurtrie, E.K., Johanningsmeier, S.D., Price, R.E., Breidt, F. 2019. Effect of brine acidification on fermentation microbiota and texture quality of cucumbers fermented in calcium chloride brines. Journal of Food Science. 84(5):1129-1137.
|
Progress 10/01/17 to 09/30/18
Outputs
Progress Report Objectives (from AD-416): 1. Determining the safety of low and alternative salt fermentations, produced nationally and internationally. 2. Develop predictive models for 5-log reduction times for pathogenic Escherichia coli in fermented and acidified vegetable products. 3. Enhance buffer capacity models for predicting pH changes in acidified foods with low acid ingredients. Approach (from AD-416): The experimental approaches that will be use to achieve the objectives will include mathematical modeling, molecular ecology studies, and biochemical analysis of fermentation brines. Specifically, for Objective 1, to determine the effects of salts on pathogen reduction in fermentations, growth and death of bacterial pathogen cocktails (strain mixtures) will be measured in fermentations by conventional bacterial plating methods using automated plating equipment. Log reduction times for pathogens will be calculated using linear or nonlinear (Weibull) models. Biochemical analysis for salts, organic acids and sugars, will be done by titration (for salts), and high performance liquid chromatography (for acids and sugars). A matrix of salt types and concentrations will be tested to determine how salt effects pathogen die-off. For Objective 2, mathematical modeling approaches to determine the reduction in pathogen populations during fermentation will utilize non-linear systems of ordinary differential equations (rate equations) using Matlab computer software. In addition computer simulation models will be developed using the C++ programming language. Data for these models will be obtained from the experiments in Objective 1. Model results will be compared to data generated under a variety of conditions to determine if the models accurately describe the data. To accomplish Objective 3, predicting pH of buffered acidified foods with low acid additives, mathematical models will be based on published ionic equilibria equations for buffered acid and base solutions. Novel methods for numerical solutions to these equations will be implemented with Matlab software. An automated titrator will be used to confirm predicted buffer capacity curve data. To fit data to the models, several optimization algorithms will be used from the Matlab Optimization Toolkit, or independently programmed in Matlab or C++. The knowledge gained will be used to help processors and regulatory agencies assess and assure the safety of acidified and fermented food products. The effects of salt on the competitive growth of pathogenic and lactic acid producing bacteria in a cucumber juice medium were determined. These data included three salt conditions: the concentration and type used in typical commercial cucumber fermentations (6% sodium chloride, NaCl), the experimental low salt cucumber fermentation currently undergoing trials at commercial fermentation plants (1.1% calcium chloride), and kimchi, sauerkraut and other ready-to-eat fermented vegetables (2% NaCl). The data show that regardless of salt concentration, the time to pathogens die off (pathogenic Escherichia coli strains) in the presence of lactic acid bacteria was similar, with 1 to 3 log reduction occurring within a 2- 3 hours window in the cucumber juice model system. These data are consistent with previous results showing that the salt type or concentration did not greatly alter the death kinetics of E. coli O157:H7 in pure culture experiments with fermentation acids. Cucumber fermentations in 1.1% calcium chloride or 6% NaCl showed similar patterns in the reduction in naturally occurring Enterobacteriacea during fermentation, demonstrating that the findings in model systems are applicable to whole cucumber fermentations. These results are significant because they contradict the widely held assumption by the food safety community that salt concentration is a primary safety factor in vegetable fermentations. Now, additional factors, such as differences in growth rates and acid resistance will have added importance and be the subject of future investigations. These data directly support research project plan to discover how different salts affect the growth and survival of pathogenic E. coli in vegetable fermentations. Development has continued on the computer simulation model for competitive growth of bacteria in vegetable fermentations. The model used realistic parameters for controlling cell growth, including sugar utilization rates (allowing cell division), and inhibition of metabolism by organic acids. The model has currently been tested using parameters derived from laboratory measurements of the biochemistry of vegetable broth (cucumber juice) fermentation by Lactobacillus plantarum, which is the most acid resistant bacterium that usually ends up dominating most vegetable fermentations. Two key parameters for the model have been identified; the per-cell rate of sugar utilization, and the amount of sugar that needs to be processed by a cell to allow cell division. These parameters were used to generate model growth curves that accurately predict the growth and acid production in laboratory fermentations. This novel cell-based simulation approach has proved to be flexible and to closely reflect the actual mechanisms controlling bacterial growth. Continued work will include incorporating data to expand the model and allow effects of bacterial competition to be predicted for complex multi- cell fermentations. Significant progress was made fitting buffer capacity models using a novel Fourier series approach to rapidly fit buffer capacity curves. Using this method, the fitted model can be further analyzed to pick out a set of buffers, based on peaks in the buffer capacity curves, that allow predicting the pH of the solution. The model was validated using acid ingredients representative of acid food products. This approach has been used to demonstrate that the pH of solutions with complex (or unknown) buffering components can be accurately predicted for both acid foods and low acid (high pH) foods that are typically ingredients in acid or acidified foods. These results build on previous results showing pH can be accurately predicted for defined mixtures of laboratory acid and base solutions. These data will aid both industry and FDA for clarifying regulatory questions about how to define acid foods (naturally below pH 4. 6) and acidified foods (high pH foods to which acid is added to bring the pH down below 4.6). Research supported by a research agreement with Pickle Packers International was done to determine whether the application of a brief blanching step could improve the safety of non-pasteurized acidified vegetable products (including refrigerated pickles). Previous research has shown that pathogenic bacteria such as Escherichia coli O157:H7 could survive for > 25% of the shelf life of refrigerated pickles, which are not heat processed. The current data support the hypothesis that blanching may be a useful method for improving safety of refrigerated pickles without negatively impacting quality. A reassessment of the cucumber fermentation microbiota using culture independent and dependent techniques was performed. The presence of opportunistic bacterial pathogens such as Citrobacter freundii, Citrobacter brakii, Enterobacter spp., Pseudomonas fluorescens, Stenotrophomonas maltophilia and Kluyvera cryocrescens, and the antibiotic resistant pathogen Serratia marcescens in the initial stage of commercial cucumber fermentations was recognized. Salt in fermentations of at least 2% salt (NaCl) or higher and decreasing pH (below pH 5.2) was inhibitory for these populations. The population density corresponding to these organisms reached undetectable levels by DNA sequence analysis in commercial cucumber fermentations by day 10, presumably due to the developing acidity from the conversion of sugars to acids. Accomplishments 01 Establishment of standards for challenge studies for processing cold- filled acidified food products. To file a scheduled process for acidified foods producers must cite or carry out a scientific study to determine if the product meets federal food safety standards. ARS researchers at Raleigh, North Carolina had a leading role in the development of a protocol (and a webinar) detailing the appropriate scientific methods for challenge studies for the assurance of safety of cold filled acidified foods that do not receive a heat process. The webinar was developed and presented in collaboration with scientists from the University of Georgia, Athens, Georgia and the University of Wisconsin, Madison, Wisconsin. The webinar was hosted by the �Beverage and Acid/Acidified Foods Professional Development Group� of the International Association for Food Protection. Over 200 people registered as attendees. This webinar is now freely available on the IAFP website. The protocol details the scientific considerations needed to conduct such a study, including the methods for growing cells, conducting an acid challenge, and analyzing the data. Many factors can influence the survival of bacterial pathogens in acid and acidified foods, and researchers may not be aware of some of them. We reviewed these factors, primarily based on previous publications from the ARS, Raleigh, North Carolina laboratory, and described how to appropriately control them. The webinar will be useful to researchers, industry stakeholders, and aid FDA to help assure that challenge studies are done with consideration of details that can assure safety. 02 Determining the presence of nitrate and nitrite in fermented and acidified vegetables. The influence of nitrate and nitrite in foods on human health has been controversial, with literature citing both positive and negative health effects. ARS researchers at Raleigh, North Carolina, measured the concentration of these compounds in a wide variety of acidified vegetables (made by adding vinegar or other acids to fresh fruits and vegetables), as well as some fermented foods currently available in the U.S. market. This was done in collaboration with a researcher from Jiangnan University, Wuxi, China, who was a visiting scientist at the Raleigh, North Carolina location. The naturally present antioxidants in foods were also of interest in the study of nitrite and nitrate levels in foods due to interactions between these compounds, so antioxidant levels were also measured. We found that nitrite was relatively rare in acidified vegetables, but was present in some of the fermented foods tested. Nitrate, on the other hand, was found to be present at varying levels in many acidified products. These results provide new information for evaluating nitrate and nitrite content in pickled fruit and vegetable products, and may be used to help assess the potential health consequences of these compounds in US consumer diets.
Impacts
(N/A)
Publications
- Ding, Z., Johanningsmeier, S.D., Price, R.E., Reynolds, R., Truong, V., Conley Payton, S.B., Breidt, F. 2018. Evaluation of nitrate and nitrite contents in pickled fruit and vegetable products. Food Control. 90:304-311.
- Kay, K., Breidt, F., Fratamico, P.M., Baranzoni, G., Kim, G., Grunden, A., Oh, D. 2017. Escherichia coli O157:H7 acid sensitivity correlates with flocculation phenotype during nutrient limitation. Frontiers in Microbiology. 8:1404.
|
Progress 10/01/16 to 09/30/17
Outputs
Progress Report Objectives (from AD-416): 1. Determining the safety of low and alternative salt fermentations, produced nationally and internationally. 2. Develop predictive models for 5-log reduction times for pathogenic Escherichia coli in fermented and acidified vegetable products. 3. Enhance buffer capacity models for predicting pH changes in acidified foods with low acid ingredients. Approach (from AD-416): The experimental approaches that will be use to achieve the objectives will include mathematical modeling, molecular ecology studies, and biochemical analysis of fermentation brines. Specifically, for Objective 1, to determine the effects of salts on pathogen reduction in fermentations, growth and death of bacterial pathogen cocktails (strain mixtures) will be measured in fermentations by conventional bacterial plating methods using automated plating equipment. Log reduction times for pathogens will be calculated using linear or nonlinear (Weibull) models. Biochemical analysis for salts, organic acids and sugars, will be done by titration (for salts), and high performance liquid chromatography (for acids and sugars). A matrix of salt types and concentrations will be tested to determine how salt effects pathogen die-off. For Objective 2, mathematical modeling approaches to determine the reduction in pathogen populations during fermentation will utilize non-linear systems of ordinary differential equations (rate equations) using Matlab computer software. In addition computer simulation models will be developed using the C++ programming language. Data for these models will be obtained from the experiments in Objective 1. Model results will be compared to data generated under a variety of conditions to determine if the models accurately describe the data. To accomplish Objective 3, predicting pH of buffered acidified foods with low acid additives, mathematical models will be based on published ionic equilibria equations for buffered acid and base solutions. Novel methods for numerical solutions to these equations will be implemented with Matlab software. An automated titrator will be used to confirm predicted buffer capacity curve data. To fit data to the models, several optimization algorithms will be used from the Matlab Optimization Toolkit, or independently programmed in Matlab or C++. The knowledge gained will be used to help processors and regulatory agencies assess and assure the safety of acidified and fermented food products. The effects of sodium chloride (NaCl) or calcium chloride concentrations on the growth and death of lactic acid bacteria (LAB, from vegetable fermentations) and pathogenic strains of Escherichia (E.) coli (STEC) were investigated. STEC strains were targeted in these studies because they are the most acid resistant pathogens in fermented and acidified foods. The data indicate that contrary to expectations, the growth of STEC strains was not significantly affected by even very high salt concentrations (to 6% NaCl, typical of commercial fermentations) in a vegetable brine, compared to a control treatment with no salt. However, the growth of LAB, which ferment vegetables in salt brines, was found to increase as salt concentrations (sodium or calcium) were increased. Data on the die-off of STEC or LAB with varying salt concentrations were also generated. In general it was determined that killing rates for LAB and STEC were influenced by the salts tested. However, calcium salt did not affect killing rates compared to NaCl for the concentrations typical of commercial cucumber fermentations. This is significant because calcium fermentation technology is being developed to help alleviate waste NaCl disposal problems at commercial pickle plants. A computer simulation model of bacterial cell growth for LAB and E. coli (STEC, defined above) under conditions typical of vegetable fermentations have been developed. Data for model validation was generated with three different salt conditions, representative of cabbage fermentations (2% NaCl), commercial cucumber fermentations (6% NaCl) and commercial cucumber fermentations brined with calcium chloride (1.1% calcium chloride). While further data is still needed to fully validate the model, the values of some model parameters, such as the rate as which sugar is used by growing cultures, and the amount of sugar needed for cells to divide have been determined. These results are important, since the outcome of competitive growth of STEC and LAB depends on sugar utilization, which in turn leads to the acid production that inhibits STEC and other pathogenic bacteria. Continued work will include the comparison of simulation results with competitive growth experiments. Once successfully validated, these models may be used to predict the die off of pathogens in competitive growth with lactic acid bacteria in vegetable fermentations. A buffer capacity model has been developed that can predict how pH changes occur with the addition of organic acids or bases to fermented or acidified vegetable brines (or dressings). Data for the model was generated using titration curves done with a strong acid (hydrochloric acid, HCl) or base (sodium hydroxide, NaOH). The model has been used to predict the pH of mixtures of organic and inorganic acids and bases. An important accomplishment was devising a means to fit the buffer capacity curve data by using a secondary model (based on trigonometric functions) that allows easy prediction of pH with the primary buffer capacity model. While further studies with foods, including fermentation brines are needed, the existing data demonstrate that pH prediction with complex mixtures of known or unknown components can be accurately done. The model may be used (possibly with further development) to determine the pH of mixtures of acidic foods with low acid ingredients (having high pH) in acidified foods. Data will be generated to determine if the pH resulting from a mixture of the food ingredients matches predictions. The model will be used to help the Food and Drug Administration (FDA) determine how to regulate the addition of small amounts of low acid (high pH) ingredients into fermented and acidified foods, without affecting safety. A buffer capacity model has been developed that can predict how pH would change with the addition of organic acids or bases to fermented or acidified vegetable brines (or dressings) by analyzing data from titration curves done with a strong acid (HCl) or base (NaOH). The model has now been used to predict the pH of mixtures of organic and inorganic acids and bases. An important accomplishment that allowed this result was devising a means to fit the buffer capacity curve data to a secondary model (based on trigonometric functions) that allows easy prediction of pH with the primary buffer capacity model. While further studies with foods, including fermentation brines, are needed, the existing data demonstrates that pH prediction with complex mixtures of known or unknown components can be accurately done. The model may be used (possibly with further development) to determine the pH of mixtures of acidic foods with low acid ingredients (having high pH) in acidified foods. Data will be generated to determine if the pH resulting from a mixture of the food ingredients matches predictions. These data will help the FDA determine how to regulate the addition of small amounts of low acid ingredients into fermented and acidified foods. Accomplishments 01 Identification of regulatory genes affecting acid resistance of pathogenic E. coli strains (STEC) in acidified vegetables. ARS scientists in Raleigh, North, Carolina, examined disease causing bacterial strains (STEC) to determine how differences in acid resistance could be explained by differences in gene regulation. From previous research, we knew that there was a wide range of acid resistance levels among these bacteria. Our research focused on determining mechanism(s) by which these important food pathogens are able to survive in acid conditions representative of acidic food products. We found that a regulatory mechanism involving small RNA molecules inside the cells apparently links expression of acid resistance genes to genes involved in producing a fibrous network that can aid STEC cells in attaching to surfaces. This regulatory network is very complex. We have found that it includes genes that are involved in the cellular response to nutrient limitation by using different kinds of growth media (with varying amounts of nutrients). The knowledge gained increases our understanding of how cells respond to acid conditions, and why some STEC strains survive better than others in acidic foods. Results may be used to help devise strategies for reducing the threat of these pathogens in acidic foods.
Impacts
(N/A)
Publications
- Baranzoni, G., Fratamico, P.M., Reichenberger, E.R., Kim, G., Breidt, F., Kay, K., Oh, D. 2016. Complete genome sequences of Escherichia coli O157:H7 strains SRCC 1675 and 28RC that vary in acid resistance. Genome Announcements. 4:4. doi: 10.1128/genomeA.00743-16.
- Kim, G., Fratamico, P.M., Breidt, F., Oh, D. 2016. Survival and expression of acid resistance genes in Shiga toxin-producing Escherichia coli acid adapted in pineapple juice and exposed to synthetic gastric fluid. Journal of Applied Microbiology. doi: 10.1111/jam.13223.
|
|